[Bug]: Llama4 Maverick runtime error (shuffle_rows) #21322
Description
Your current environment
The output of python collect_env.py
==============================
System Info
==============================
OS : CentOS Stream 9 (x86_64)
GCC version : (GCC) 11.5.0 20240719 (Red Hat 11.5.0-7)
Clang version : Could not collect
CMake version : version 4.0.3
Libc version : glibc-2.34
==============================
PyTorch Info
==============================
PyTorch version : 2.7.1+cu128
Is debug build : False
CUDA used to build PyTorch : 12.8
ROCM used to build PyTorch : N/A
==============================
Python Environment
==============================
Python version : 3.12.10 (main, May 9 2025, 00:00:00) [GCC 11.5.0 20240719 (Red Hat 11.5.0-5)] (64-bit runtime)
Python platform : Linux-6.4.3-0_fbk20_zion_2830_g3e5ab162667d-x86_64-with-glibc2.34
==============================
CUDA / GPU Info
==============================
Is CUDA available : True
CUDA runtime version : 12.8.93
CUDA_MODULE_LOADING set to : LAZY
GPU models and configuration :
GPU 0: NVIDIA H100
GPU 1: NVIDIA H100
GPU 2: NVIDIA H100
GPU 3: NVIDIA H100
GPU 4: NVIDIA H100
GPU 5: NVIDIA H100
GPU 6: NVIDIA H100
GPU 7: NVIDIA H100
Nvidia driver version : 535.154.05
cuDNN version : Probably one of the following:
/usr/lib64/libcudnn.so.9.5.1
/usr/lib64/libcudnn_adv.so.9.5.1
/usr/lib64/libcudnn_cnn.so.9.5.1
/usr/lib64/libcudnn_engines_precompiled.so.9.5.1
/usr/lib64/libcudnn_engines_runtime_compiled.so.9.5.1
/usr/lib64/libcudnn_graph.so.9.5.1
/usr/lib64/libcudnn_heuristic.so.9.5.1
/usr/lib64/libcudnn_ops.so.9.5.1
HIP runtime version : N/A
MIOpen runtime version : N/A
Is XNNPACK available : True
==============================
CPU Info
==============================
Architecture: x86_64
CPU op-mode(s): 32-bit, 64-bit
Address sizes: 52 bits physical, 57 bits virtual
Byte Order: Little Endian
CPU(s): 384
On-line CPU(s) list: 0-383
Vendor ID: AuthenticAMD
Model name: AMD EPYC 9654 96-Core Processor
CPU family: 25
Model: 17
Thread(s) per core: 2
Core(s) per socket: 96
Socket(s): 2
Stepping: 1
Frequency boost: enabled
CPU(s) scaling MHz: 83%
CPU max MHz: 3707.8120
CPU min MHz: 1500.0000
BogoMIPS: 4792.60
Flags: fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pat pse36 clflush mmx fxsr sse sse2 ht syscall nx mmxext fxsr_opt pdpe1gb rdtscp lm constant_tsc rep_good amd_lbr_v2 nopl nonstop_tsc cpuid extd_apicid aperfmperf rapl pni pclmulqdq monitor ssse3 fma cx16 pcid sse4_1 sse4_2 x2apic movbe popcnt aes xsave avx f16c rdrand lahf_lm cmp_legacy svm extapic cr8_legacy abm sse4a misalignsse 3dnowprefetch osvw ibs skinit wdt tce topoext perfctr_core perfctr_nb bpext perfctr_llc mwaitx cpb cat_l3 cdp_l3 invpcid_single hw_pstate ssbd mba perfmon_v2 ibrs ibpb stibp ibrs_enhanced vmmcall fsgsbase bmi1 avx2 smep bmi2 erms invpcid cqm rdt_a avx512f avx512dq rdseed adx smap avx512ifma clflushopt clwb avx512cd sha_ni avx512bw avx512vl xsaveopt xsavec xgetbv1 xsaves cqm_llc cqm_occup_llc cqm_mbm_total cqm_mbm_local avx512_bf16 clzero irperf xsaveerptr rdpru wbnoinvd amd_ppin cppc arat npt lbrv svm_lock nrip_save tsc_scale vmcb_clean flushbyasid decodeassists pausefilter pfthreshold avic v_vmsave_vmload vgif x2avic v_spec_ctrl vnmi avx512vbmi umip pku ospke avx512_vbmi2 gfni vaes vpclmulqdq avx512_vnni avx512_bitalg avx512_vpopcntdq rdpid overflow_recov succor smca fsrm flush_l1d
Virtualization: AMD-V
L1d cache: 6 MiB (192 instances)
L1i cache: 6 MiB (192 instances)
L2 cache: 192 MiB (192 instances)
L3 cache: 768 MiB (24 instances)
NUMA node(s): 2
NUMA node0 CPU(s): 0-95,192-287
NUMA node1 CPU(s): 96-191,288-383
Vulnerability Gather data sampling: Not affected
Vulnerability Itlb multihit: Not affected
Vulnerability L1tf: Not affected
Vulnerability Mds: Not affected
Vulnerability Meltdown: Not affected
Vulnerability Mmio stale data: Not affected
Vulnerability Retbleed: Not affected
Vulnerability Spec store bypass: Mitigation; Speculative Store Bypass disabled via prctl
Vulnerability Spectre v1: Mitigation; usercopy/swapgs barriers and __user pointer sanitization
Vulnerability Spectre v2: Vulnerable: eIBRS with unprivileged eBPF
Vulnerability Srbds: Not affected
Vulnerability Tsx async abort: Not affected
==============================
Versions of relevant libraries
==============================
[pip3] mypy-extensions==1.0.0
[pip3] numpy==1.26.4
[pip3] nvidia-cublas-cu12==12.8.3.14
[pip3] nvidia-cuda-cupti-cu12==12.8.57
[pip3] nvidia-cuda-nvrtc-cu12==12.8.61
[pip3] nvidia-cuda-runtime-cu12==12.8.57
[pip3] nvidia-cudnn-cu12==9.7.1.26
[pip3] nvidia-cufft-cu12==11.3.3.41
[pip3] nvidia-cufile-cu12==1.13.0.11
[pip3] nvidia-curand-cu12==10.3.9.55
[pip3] nvidia-cusolver-cu12==11.7.2.55
[pip3] nvidia-cusparse-cu12==12.5.7.53
[pip3] nvidia-cusparselt-cu12==0.6.3
[pip3] nvidia-nccl-cu12==2.26.2
[pip3] nvidia-nvjitlink-cu12==12.8.61
[pip3] nvidia-nvtx-cu12==12.8.55
[pip3] pyzmq==27.0.0
[pip3] sentence-transformers==3.2.1
[pip3] torch==2.7.1+cu128
[pip3] torchaudio==2.7.1+cu128
[pip3] torchvision==0.22.1+cu128
[pip3] transformers==4.53.2
[pip3] transformers-stream-generator==0.0.5
[pip3] triton==3.3.1
[pip3] tritonclient==2.51.0
[pip3] vector-quantize-pytorch==1.21.2
[conda] Could not collect
==============================
vLLM Info
==============================
ROCM Version : Could not collect
Neuron SDK Version : N/A
vLLM Version : 0.9.2rc2.dev304+g28a6d5423 (git sha: 28a6d5423)
vLLM Build Flags:
CUDA Archs: Not Set; ROCm: Disabled; Neuron: Disabled
GPU Topology:
GPU0 GPU1 GPU2 GPU3 GPU4 GPU5 GPU6 GPU7 NIC0 NIC1 NIC2 NIC3 CPU Affinity NUMA Affinity GPU NUMA ID
GPU0 X NV18 NV18 NV18 NV18 NV18 NV18 NV18 SYS SYS SYS SYS 0-95,192-287 0 N/A
GPU1 NV18 X NV18 NV18 NV18 NV18 NV18 NV18 PHB PHB SYS SYS 0-95,192-287 0 N/A
GPU2 NV18 NV18 X NV18 NV18 NV18 NV18 NV18 SYS SYS SYS SYS 0-95,192-287 0 N/A
GPU3 NV18 NV18 NV18 X NV18 NV18 NV18 NV18 SYS SYS SYS SYS 0-95,192-287 0 N/A
GPU4 NV18 NV18 NV18 NV18 X NV18 NV18 NV18 SYS SYS SYS SYS 96-191,288-383 1 N/A
GPU5 NV18 NV18 NV18 NV18 NV18 X NV18 NV18 SYS SYS SYS SYS 96-191,288-383 1 N/A
GPU6 NV18 NV18 NV18 NV18 NV18 NV18 X NV18 SYS SYS PHB PHB 96-191,288-383 1 N/A
GPU7 NV18 NV18 NV18 NV18 NV18 NV18 NV18 X SYS SYS SYS SYS 96-191,288-383 1 N/A
NIC0 SYS PHB SYS SYS SYS SYS SYS SYS X PIX SYS SYS
NIC1 SYS PHB SYS SYS SYS SYS SYS SYS PIX X SYS SYS
NIC2 SYS SYS SYS SYS SYS SYS PHB SYS SYS SYS X PIX
NIC3 SYS SYS SYS SYS SYS SYS PHB SYS SYS SYS PIX X
Legend:
X = Self
SYS = Connection traversing PCIe as well as the SMP interconnect between NUMA nodes (e.g., QPI/UPI)
NODE = Connection traversing PCIe as well as the interconnect between PCIe Host Bridges within a NUMA node
PHB = Connection traversing PCIe as well as a PCIe Host Bridge (typically the CPU)
PXB = Connection traversing multiple PCIe bridges (without traversing the PCIe Host Bridge)
PIX = Connection traversing at most a single PCIe bridge
NV# = Connection traversing a bonded set of # NVLinks
NIC Legend:
NIC0: mlx5_0
NIC1: mlx5_1
NIC2: mlx5_2
NIC3: mlx5_3
==============================
Environment Variables
==============================
CUDA_CACHE_PATH=/data/users/yming/.nv/ComputeCache
LD_LIBRARY_PATH=/usr/local/cuda-12.8/lib64/:/usr/local/cuda-12.8/lib64/:
NCCL_CUMEM_ENABLE=0
PYTORCH_NVML_BASED_CUDA_CHECK=1
TORCHINDUCTOR_COMPILE_THREADS=1
CUDA_MODULE_LOADING=LAZY```
🐛 Describe the bug
llama4 maverick is failing to start due to the runtime error during shuffle_row.
this can be reproduced:
vllm serve meta-llama/Llama-4-Maverick-17B-128E-Instruct-FP8 --max_model_len 8192 --kv_cache_dtype fp8 --enable-expert-parallel --tensor-parallel-size 8 --trust-remote-code --gpu-memory-utilization 0.8 --disable-log-requests
This is likely related to #20762 @ElizaWszola
this can also be reproduced with pytest -s tests/models/multimodal/generation/test_maverick.py, which requires only 2xH100 by running dummy version of maverick.
cc @yeqcharlotte @luccafong @houseroad
(VllmWorker rank=0 pid=938767) ERROR 07-21 11:16:11 [multiproc_executor.py:546] File "/data/users/yming/gitrepos/vllm/vllm/model_executor/layers/fused_moe/layer.py", line 1579, in moe_forward
(VllmWorker rank=0 pid=938767) ERROR 07-21 11:16:11 [multiproc_executor.py:546] return self.forward_impl(hidden_states, router_logits)
(VllmWorker rank=0 pid=938767) ERROR 07-21 11:16:11 [multiproc_executor.py:546] ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
(VllmWorker rank=0 pid=938767) ERROR 07-21 11:16:11 [multiproc_executor.py:546] File "/data/users/yming/gitrepos/vllm/vllm/model_executor/layers/fused_moe/layer.py", line 1489, in forward_impl
(VllmWorker rank=0 pid=938767) ERROR 07-21 11:16:11 [multiproc_executor.py:546] final_hidden_states = self.quant_method.apply(
(VllmWorker rank=0 pid=938767) ERROR 07-21 11:16:11 [multiproc_executor.py:546] ^^^^^^^^^^^^^^^^^^^^^^^^
(VllmWorker rank=0 pid=938767) ERROR 07-21 11:16:11 [multiproc_executor.py:546] File "/data/users/yming/gitrepos/vllm/vllm/model_executor/layers/quantization/compressed_tensors/compressed_tensors_moe.py", line 959, in apply
(VllmWorker rank=0 pid=938767) ERROR 07-21 11:16:11 [multiproc_executor.py:546] return cutlass_moe_fp8(
(VllmWorker rank=0 pid=938767) ERROR 07-21 11:16:11 [multiproc_executor.py:546] ^^^^^^^^^^^^^^^^
(VllmWorker rank=0 pid=938767) ERROR 07-21 11:16:11 [multiproc_executor.py:546] File "/data/users/yming/gitrepos/vllm/vllm/model_executor/layers/fused_moe/cutlass_moe.py", line 414, in cutlass_moe_fp8
(VllmWorker rank=0 pid=938767) ERROR 07-21 11:16:11 [multiproc_executor.py:546] return fn(
(VllmWorker rank=0 pid=938767) ERROR 07-21 11:16:11 [multiproc_executor.py:546] ^^^
(VllmWorker rank=0 pid=938767) ERROR 07-21 11:16:11 [multiproc_executor.py:546] File "/home/yming/uv_env/vllm/lib64/python3.12/site-packages/torch/nn/modules/module.py", line 1751, in _wrapped_call_impl
(VllmWorker rank=0 pid=938767) ERROR 07-21 11:16:11 [multiproc_executor.py:546] return self._call_impl(*args, **kwargs)
(VllmWorker rank=0 pid=938767) ERROR 07-21 11:16:11 [multiproc_executor.py:546] ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
(VllmWorker rank=0 pid=938767) ERROR 07-21 11:16:11 [multiproc_executor.py:546] File "/home/yming/uv_env/vllm/lib64/python3.12/site-packages/torch/nn/modules/module.py", line 1762, in _call_impl
(VllmWorker rank=0 pid=938767) ERROR 07-21 11:16:11 [multiproc_executor.py:546] return forward_call(*args, **kwargs)
(VllmWorker rank=0 pid=938767) ERROR 07-21 11:16:11 [multiproc_executor.py:546] ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
(VllmWorker rank=0 pid=938767) ERROR 07-21 11:16:11 [multiproc_executor.py:546] File "/data/users/yming/gitrepos/vllm/vllm/model_executor/layers/fused_moe/modular_kernel.py", line 770, in forward
(VllmWorker rank=0 pid=938767) ERROR 07-21 11:16:11 [multiproc_executor.py:546] fused_out = self._maybe_chunk_fused_experts(
(VllmWorker rank=0 pid=938767) ERROR 07-21 11:16:11 [multiproc_executor.py:546] ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
(VllmWorker rank=0 pid=938767) ERROR 07-21 11:16:11 [multiproc_executor.py:546] File "/data/users/yming/gitrepos/vllm/vllm/model_executor/layers/fused_moe/modular_kernel.py", line 545, in _maybe_chunk_fused_experts
(VllmWorker rank=0 pid=938767) ERROR 07-21 11:16:11 [multiproc_executor.py:546] return self._do_fused_experts(
(VllmWorker rank=0 pid=938767) ERROR 07-21 11:16:11 [multiproc_executor.py:546] ^^^^^^^^^^^^^^^^^^^^^^^
(VllmWorker rank=0 pid=938767) ERROR 07-21 11:16:11 [multiproc_executor.py:546] File "/data/users/yming/gitrepos/vllm/vllm/model_executor/layers/fused_moe/modular_kernel.py", line 492, in _do_fused_experts
(VllmWorker rank=0 pid=938767) ERROR 07-21 11:16:11 [multiproc_executor.py:546] self.fused_experts.apply(
(VllmWorker rank=0 pid=938767) ERROR 07-21 11:16:11 [multiproc_executor.py:546] File "/data/users/yming/gitrepos/vllm/vllm/model_executor/layers/fused_moe/cutlass_moe.py", line 314, in apply
(VllmWorker rank=0 pid=938767) ERROR 07-21 11:16:11 [multiproc_executor.py:546] run_cutlass_moe_fp8(
(VllmWorker rank=0 pid=938767) ERROR 07-21 11:16:11 [multiproc_executor.py:546] File "/data/users/yming/gitrepos/vllm/vllm/model_executor/layers/fused_moe/cutlass_moe.py", line 160, in run_cutlass_moe_fp8
(VllmWorker rank=0 pid=938767) ERROR 07-21 11:16:11 [multiproc_executor.py:546] a1q_scale = (ops.shuffle_rows(a1q_scale, a_map)
(VllmWorker rank=0 pid=938767) ERROR 07-21 11:16:11 [multiproc_executor.py:546] ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
(VllmWorker rank=0 pid=938767) ERROR 07-21 11:16:11 [multiproc_executor.py:546] File "/data/users/yming/gitrepos/vllm/vllm/_custom_ops.py", line 908, in shuffle_rows
(VllmWorker rank=0 pid=938767) ERROR 07-21 11:16:11 [multiproc_executor.py:546] torch.ops._moe_C.shuffle_rows(input_tensor, dst2src_map, output_tensor)
(VllmWorker rank=0 pid=938767) ERROR 07-21 11:16:11 [multiproc_executor.py:546] File "/home/yming/uv_env/vllm/lib64/python3.12/site-packages/torch/_ops.py", line 1158, in __call__
(VllmWorker rank=0 pid=938767) ERROR 07-21 11:16:11 [multiproc_executor.py:546] return self._op(*args, **(kwargs or {}))
(VllmWorker rank=0 pid=938767) ERROR 07-21 11:16:11 [multiproc_executor.py:546] ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
(VllmWorker rank=0 pid=938767) ERROR 07-21 11:16:11 [multiproc_executor.py:546] RuntimeError: num_cols must be divisible by 128 / sizeof(input_tensor.scalar_type()) / 8
Before submitting a new issue...
- Make sure you already searched for relevant issues, and asked the chatbot living at the bottom right corner of the documentation page, which can answer lots of frequently asked questions.